JPS583471A - Picture processing method - Google Patents

Abstract

PURPOSE:To perform highly efficient transmission possible for picture reproduction with high quality, by determining the share of density of small picture elements with the state of eight adjacent original picture elements, for the picture elements of intermediate density out of picture elements signals encoded and transmitted in 3 levels of white, black and intermediate density. CONSTITUTION:A reading transmitter 30 transmits readout picture element signals to a reception side via a transmission line after encoding and quantizing in 3 stages of 0, 1 and 2. A signal decoded at a decoder 32 at the reception side is transferred to a buffer memory 33, signals of eight picture elements adjacent to a remarked picture element are picked up from each terminal and the eight picture elements are added at eight adders 34A-34H. The output sum T1-T8 of the adjacent eight picture elements of each adder are respectively inputted to a maximum value detection circuit 35 to detect a maximum value TM. Depending on the TM belonging to the sum of 8 picture elements, i.e., either of the T1-T8, the value of the small picture elements sectioning the remarked picture element into small sections can be determined.

Description

DETAILED DESCRIPTION OF THE INVENTION The present invention relates to an image processing method, and more particularly to an image processing method that enables low-density transmission and high-density reproduction.

Conventionally, for the purpose of efficient use of the transmission system and high-speed transmission, and also for the reproduction of i-high quality images on the reproduction side,
A method has been adopted in which images are transmitted at low density and reproduced at high density.

As a method of this kind, a method is conventionally known in which one image is transmitted to each pixel, and then the image is transferred to one pixel on the receiving side and then binary-restored as four small pixels.

However, in this case, since no suitable restoration method has been found, it is necessary to transmit five density levels corresponding to the case where all four small pixels are white to the case where all four small pixels are black. (For example, JP-A-53-138
No. 630) In order to transmit this type of information in binary, K requires at least a 3-bit signal, but since 8 pieces of information can be transmitted with 3 bits, this is a desirable state in terms of transmission efficiency. It wasn't.

This invention was made based on the above circumstances,
Based on a new restoration method! An object of the present invention is to provide an image processing method that can achieve high-efficiency transmission capable of high-quality image reproduction at a ternary level without requiring 5-level transmission.

In order to achieve this object, according to the present invention, the density fV of the pixel read in pixel units is classified into three levels of white, black, and intermediate density and encoded and transmitted, and on the receiving side, each pixel is divided into a plurality of small In an image processing method that divides pixels into pixels and allocates white and black densities to these subpixels, if the received density of a pixel of interest is white or black, the subpixels corresponding to this pixel of interest are divided into all white or black pixels, respectively. If the received density of the pixel of interest is an intermediate density, the density distribution of each subpixel corresponding to the pixel of interest is determined based on the states of eight original pixels adjacent to the pixel of interest.

The present invention will be described in detail below.

On the image reading and transmitting side, the density of the image is read for each pixel, and the density level is quantized into three levels: 0 (white), 1 (intermediate density = gray), and 2 (black).

This quantized signal is encoded, for example, as shown in Table 1 below, and is transmitted with a run length code as shown in Table 2.

Table 1 Table 2 On the receiving side, the received pixel of interest E"i as shown in FIG.
a - 8 centrally adjacent pixels A, B, C;

Taking into consideration D, F, G, H, and I, the pixel EV is divided into small pixels a, b, e, and d as shown in Figure 2 and output.0 of each small pixel a% d shown in the second factor Densities 0 (white) and 1 (black) are determined as follows: 0 (1) E=O
When, a = b = c = d (2) When E = 2, a = b 4 c: d =, 1 (3) When E = 1, TI = A + B + D T2 = A + B + C T3 = B + C + F T4 = A + D + G Let T5=D+G+H T6=C+F+I T7=G+H+I T8=F+H+1 and find the moth-sized TM among Tl to T8.

That is, find TM=Max(TI, T2...., T8),
The density of a small pixel is determined as follows.

(3-1) When TM=72, a=b=1. e = = d = 0 (3-2) When TM = 74, a = c = 1, b = d = 0 (3-3) TM
When = T6, b = d = 1, a = c = 0 (3-4) TM = T When 7, (B = = d = 1, a = b = 0 (3-5) TM
= TI, if TI (H, a = 1, b = e =
If d = OTl>H, a=b=c=1, d=
0(3-6) When TM=T 5, T5 (If H, c=1, a=b=d=OT5
>H, then a=e=d=1. b=0(3-7)TM
= When T8, T8 (if H, d=1, 1=i) = 6=QT8>
If H, then b=c=d=1. a=0 However, H is a threshold value, and it is desirable that H=4 or so.

Figure 3 shows an example of a configuration for executing the above processing.
It's from F. ICJ in the same figure: Bara reading deck transmitter (9)
, a transmission path 31, a decoder 32, a buffer memory O, an adder 34A to 34H, a maximum value detection circuit, a comparator A, and a ROM 37.

The reading and transmitting device (9) reads the document pixel by pixel using a scanner (not shown) or the like, and receives the read image signals vO,
The signal is quantized into three stages of 1 and 2, then encoded and sent to the receiving side via the transmission line 31.

The decoder 32 decodes the encoded image signal and transfers it to a buffer memory.

The backup memory stores, for example, three scanning lines worth of image information so that the processing described above is possible.This buffer memory stores three lines of image information as shown in Figure 1. Necessary information from a group of pixels in three columns is emitted from each of the eight terminals at a predetermined timing.

The information retrieved from the buffer memory is added in eight end adders 34A-34)I.

These adders 34A to 341 are connected to T1 to T8, respectively.
Calculate. However, from those eight terminals of the backup memory, the outputs T1 to T8 of each adder 34A to 34H are input to the maximum value detection circuit 35, and TM is detected and output.

The maximum value TM thus detected is TM TI.

T5. If it is either T8, the comparator ad threshold H
TM is compared with T2. T3. T4. If it is either T6゜TI, it passes through the comparator 36 and is stored in the ROM 37.
A signal is given to

The ROM 37 stores in advance the signal TM or the comparison result of the comparator as an address signal '*b*e.

Output the combination of values of d. In this way, the values of the small pixels 1 to d for each pixel of interest E are determined, and the output image signal S8
is sent.

By configuring as described above, this invention makes it possible to perform efficient transmission using ternary values, and also provides an image processing method that can obtain high-quality images even when restoring from ternary values. .

[Brief explanation of the drawing]

FIG. 1 is a pixel arrangement diagram of a reading 9 and a received image according to the present invention, FIG. WX2 is an arrangement diagram of small pixels on the receiving side according to the present invention, and FIG. 3 is a system diagram of an embodiment of the present invention.

Claims (1)

[Claims] In a reception processing method in which the density of a pixel read pixel by pixel is classified into three levels of white, black, and intermediate density and encoded and transmitted, when the reception density of the pixel of interest is white or black. In this case, all of the small pixels corresponding to this pixel of interest are set to white or black, and if the received density of the pixel of interest is intermediate density, each of the small pixels corresponding to this pixel of interest is An image processing method that determines the density distribution of small pixels.